CANDU & PWR
CANDU & PWR
2008 January
CANDU Reactor
2008 January
2008 January
2008 January
Administration and maintenance facilities Pump house Reactor containment Turbine and generator
2008 January
No separation of coolant from moderator Large, more complex fuel assembly Batch (off-power) refuelling Boron/chemical reactor control in coolant system
6
2008 January
2008 January
2008 January
Reactivity Devices
In CANDU: Devices are in benign environment (moderator at low pressure and temperature) Pressure-driven ejection not possible Separate devices for control and safety Modest reactivity worth Maximum total reactivity rate <0.35 mk/s
In PWR: Device worth is very high, to match high core excess reactivity Pressure-driven ejection must be considered in safety analysis Same for accidental boron dilution
2008 January
11
Reactivity Transients
A) Loss of Regulation
CANDU
2 Yes Low
PWR
1 No High
Shutdown Systems Shutdown Systems completely independent from RRS Reactivity-device worth
2008 January
12
CANDU No effect
C) Other Transients
CANDU
Not possible NA
PWR
Large positive reactivity Large positive reactivity
13
CANDU
PWR
Void Reactivity
In CANDU, Large Loss of Coolant (LLOCA) is the accident which is the most challenging in terms of positive reactivity insertion. PWR lattice has very high negative fuel-temperature (Doppler) and power coefficients, which cater to device ejection and short promptneutron lifetime. In CANDU, the fuel-temperature and power coefficients are much less negative, but the transients are generally milder and slower.
2008 January 14
Arranging heat-transport system to minimize rate of reactivity insertion on coolant voiding (e.g., subdividing the heat-transport system into 2 loops). Providing two fully capable Shutdown Systems that can individually overtake any reactivity transient.
2008 January
15
Core-Region Decoupling
The CANDU core is more decoupled than a PWR core. This means that core regions or zones can behave somewhat independently of others to a greater degree in CANDU than in PWR: the spatial power distribution can be more easily tilted. Also, refuelling occurs daily, in various core regions. A spatial-control system is more necessary in CANDU.
2008 January
16
Fuel-Cycle Safety
Natural uranium or other low-fissile-content fuel ensures that there is no potential for criticality of new or used fuel in air or light water. No need to ship new fuel in borated steel containers No need to borate the ECC System water No need to borate the fuel-bay water Simplified irradiated-fuel dry storage
2008 January
17
Reactivity devices in cool, low-pressure moderator. Rod ejection not possible. Small core excess reactivity, because of on-power refuelling. Worth of reactivity devices in RRS is low, magnitude of reactivity-induced transients is limited. Reactivity-device worth constant over life of plant. Long prompt-neutron lifetime slows rate of transients. Nuclear lattice (lattice pitch) nearly optimized for maximum reactivity. Any event that relocates the fuel reduces reactivity.
contd
2008 January 18
No reactivity effect from many postulated transients, including rapid cool-down of the heat-transport system. Moderator system can remove decay heat under such severe conditions as a LLOCA coincident with ECC failure. Low radiation fields in coolant, because of on-line failed-fuel detection and removal, and absence of chemicals for reactivity control. Easy handling of new and irradiated fuel. No criticality concern, in ordinary water or air, regardless of storage configuration. Large moderator volume serves as excellent heat sink in hypothetical severe accidents.
2008 January 19
END
2008 January
20